Testing Methodology

System Setup


Here is the system used for writing reviews such as this. You will see that we run the processor at regular and at overclocked speeds, so we can generate more heat to test TIM such as our test subject today: OCZ Freeze

• Processor: Intel Core 2 Duo E6850 @ 3.0Ghz running 1.3500V(Stock)
• Processor: Intel Core 2 Duo E6850 @ 3.6Ghz running 1.4250V(20% OC)
• CPU Cooling: Noctua NH-U12F with Noctua NF-S12-1200 Fan
• Memory: 2GB OCZ Platinum Rev. 2 DDR2 @ 900Mhz (4-4-4-12)
• Motherboard: ASUS P5N32-E SLI (680i)
• Disk Drive: ASUS DVD DRW-2014L1T
• Hard Drive: Seagate Barracuda 250GB SATAII
• OS: Windows Vista Ultimate x32
• Graphics Cards: BFG Tech 8800GTS 640Mb OC (550/1300/800MHz Stock OC)
• Drivers: Nvidia 169.28
• Monitor: Acer AL2216W (1680X1050)

The Intel Core 2 Duo E6850 is one of the fastest dual core processors to date running at 3.0GHz using only 1.35V. It runs fairly cool, however when it is overclocked, it can get pretty hot. Note that our case is open and on its side during our testing periods. We do this to decrease the cases cooling capacity, which could possibly lower the CPU's temperature and in addition, it allows us to better control the ambient temperature so it is the same during each test.

Curing times used:

- OCZ Freeze: None
- AC MX-2: None
- AS Ceramique: 25 hours
- OCZ Ultra 5+: 200 hours

OCZ Freeze does NOT have a curing time, which means testing can start right away. In our previous review, we had to wait 25 hours for the curing time for Ceramique but also OCZ Ultra 5+ which takes a mind-blowing 150-200 hours. Today we use the same recordings that we took in our last test and thus can we compare OCZ Freeze against other thermal compounds in the market. As you will notice, the temperature in the room is carefully monitored as is humidity.

Temperature Logging & Load Conditions

Stress Program: Orthos
Temperature logging program: CoreTemp 0.96.1
Ambient room temperature: 23.5°C (+/- 0.5°C)

For temperature logging, we used CoreTemp 0.96.1 and let it log the temperatures for the entire test period. After idling the computer for 20 minutes, we run Orthos to stress the two cores in our CPU for 20 minutes and then we turn our computer off to let it cool down again for 60 minutes. We take this long to let the ambient (room) temperatures settle as well. After doing all of that, we remove the heatsink, reapply the thermal compound we are testing and begin again after (if necessary) the curing time has expired. We do this for three mounts for all three compounds we are using in this review.

Why three mounts?

While this whole business of three separate mounts for three separate compounds may be a time-consuming process, there is a method to our madness. The issue with thermal compound testing is that there are so many variables to take into account and usually such a small difference between temperatures which could all be chalked up to different heatsink mounting methods from one test to the next. Those of you who are well-versed in water cooling know about the temperature changes that can be experienced from one mount to the next and the same thing goes for thermal compound.

To run one test with each compound and arbitrarily pull a winner out of our butts would not serve you much good. So, we will remount the heatsink three times for each thermal compound with a new application of TIM between each test. This way you will be able to see not only how much temperatures can change based on installation but also determine a clear-cut winner.

Of Average Temperatures...

Another one of the variables we wanted to eliminate with this testing methodology is the reading of "maximum" temperature results over a period of time. The issue with reporting maximum temperatures is that temperatures change on a millisecond basis and while a maximum temperature may be picked up by a logging program fine in one run, the next run may miss it entirely.

Instead what we are doing in this review is reporting to you the average temperature seen during both the idle and load tests. We feel that this will show a much more accurate representation as to what a thermal compound is capable of.

The Competitors


Pictured above are the two thermal compounds made by OCZ. We are also including our results from champions of past and present times; OCZ Ultra 5+ (which has now been replaced in OCZ's stable with the Freeze), Arctic Silver Ceramique and Arctic Cooling MX-2. Let’s go see if our new OCZ Freeze can take on this stiff competition.